This is a DANGER ZONE that should be avoided right now, IMHO.
Look at the world today: we have this political system that has gone haywire and media outlets that hand anybody a microphone and five minutes to spout off anything they want without fact-checking it or even taking any responsibility for its accuracy or relevance. If someone wants to talk about how they have found “reliable sources” who claim the moon is made of green cheese, there are plenty of media outlets that are happy to tive them air time, since anything controversial “sells” and boosts their ad revenues.
They feel their only obligation is to give five minutes of air time to someone at NASA or some PhD at a university who might try to debunk it. But at the end of the day, they end up creating a lot of confusion because they themselves are NOT “experts” in anything they cover, and they’ve lost their journalistic compass that tells them how to deal with this stupidity. Walter Conkite is turning in his grave.
AI is in no better of a position than today’s media outlets — you can ask it a question and it will offer up answers, but it has no idea if the answers it spits out are accurate or even relevant. And anybody who’s NOT a bona fide “expert” in the subject will have no frigging clue or any way to tell. Who ya gonna trust?
“Hey, Fox just ran a 15-minute segment where they had someone talk about how the moon really IS made of green cheese! The two peole they interviewed were very convincing, and the guys from NASA made no sense at all. So I’m inclined to believe that it really IS made of green cheese now.”
Welcome to the world of AI, where most people refuse to even question nonsense being given air time by major media outlets as if it’s real, legitimate, factual information.
The sad fact of the matter is that after enough publicity promoting stories that assert the moon IS made of green cheese get published or aired (even just online), guess what will happen? ChatGPT will start generating answers supporting that totally bogus viewpoint and people will start believing it even more.
If you’re not enough of an expert to know it’s giving you bogus info, then DO NOT TRUST IT!
It’s that simple.
Believe me when I say there WILL be folks who will start feeding these AI systems nonsense like adding arsenic to brownies just to see what happens, and it will start reporting brownie recipies that require arsenic — along with instructions on how to extract it from rat poison — and people who don’t know any better WILL TRUST IT. :-O
Then Fox will do a segment where they show that ChatGPT is offering up brownie recipies that contain arsenic and warnings from a couple of doctors who say it’ll kill you. And then they’ll inteview someone who will claim that arsenic in small doses is actually an aphrodesiac and improves your mental health and let YOU decide. They call that “balanced reporting” today.
-David Schwartz
> On Jun 6, 2023, at 1:49 PM, James Mcphee via PLUG-discuss <plug-discuss@lists.phxlinux.org> wrote:
>
> When I'm actually an expert at the thing I ask chatGPT for, yeah, like an intern. It'll say something that will prompt me to go down some road or other, and I ignore the obviously wrong answers. When I'm an amateur at the thing, it sounds authoritative and I don't have the ability to know better. But then, we're not paying for it to be RIGHT. It's still a solution we're trying to find a problem for.
---------------------------------------------------
PLUG-discuss mailing list:
PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/plug-discuss
(text/plain)